Intrinsic and extrinsic regulation of signaling for growth and guidance

Dissertation zur Erlangung des Doktorgrades Der Naturwissenschaften

an der Fakultät für Biologie der Ludwig-Maximilians-Universität München

Graziana Gatto

Intrinsic and extrinsic regulation of signaling for axon growth and guidance

Dissertation zur Erlangung des Doktorgrades Der Naturwissenschaften

der Fakultät für Biologie der Ludwig-Maximilians-Universität München

Graziana Gatto

1. Gutachter: Prof. Dr. Rüdiger Klein 2. Gutachter: Prof. Dr. Barbara Conradt

Tag der Einreichung: 16.04.2013

Tag der mündlichen Prüfung: 18.09.2013

The work presented in this dissertation was performed in the laboratory of Prof. Dr. Rüdiger Klein, Department of Molecules – Signaling – Development, Max-Planck- Institute of Neurobiology, Martinsried, Germany.

Eidesstattliche Erklärung

Ich versichere hiermit an Eides Statt, dass die vorgelegte Dissertation von mir selbständig

und ohne unerlaubte Hilfe angefertigt ist.

München, den 16.04.2013 Graziana Gatto

Erklärung

Hiermit erkläre ich, dass die Dissertation nicht ganz oder in wesentlichen Teilen einer

anderen Prüfungskommission vorgelegt worden ist und ich mich anderweitig einer

Doktorprüfung ohne Erfolg nicht unterzogen habe.

München, den 16.04.2013 Graziana Gatto

Publications from the work presented in this dissertation

Gatto G, Dudanova I, Suetterlin P, Davies AM, Drescher U, Bixby JL, Klein R

Protein Tyrosine Phosphatase Receptor Type O Inhibits Trigeminal Axon Growth and Branching by Repressing TrkB and Ret Signaling

J.Neurosci 2013 March 20; 33(12):5399-410

Dudanova I, Gatto G, Klein R

GDNF acts as a chemoattractant to support ephrinA-induced repulsion of limb motor

Current Biology 2010 December 7; 20(23): 2150-6

To my dear grandmother Graziella in loving memory

"It is the rule rather than the exception

in research that new leads come from accidental findings and these leads,

when followed, would channel

the investigation into a new direction."

(cit. Rita Levi-Montalcini)

Acknowledgements

First I would like to thank my supervisor Ruediger Klein for giving me the opportunity to work in such a great scientific environment, and for giving me the freedom to develop my own ideas. Thank you for your constant and unreserved support and for helping me to mature as a scientist. Ringrazio i miei genitori e le mie nonne per il sostegno e la fiducia che mi hanno dato e continueranno a darmi, e soprattutto perché tutto ciò che ho e che sono diventata lo devo a loro. I am grateful to Alun Davies for kindly teaching me how to culture trigeminal , and to Philipp Suetterlin and Uwe Drescher for performing the retinotopic tracings. I thank the members of my thesis advisory committee: Andrea Huber and Takashi Suzuki, for their support, their intellectual input and their continuous interest in my work. I would like to thank Irina Dudanova for introducing me to the motor world patiently teaching me all the related techniques, and for our great collaboration over the years; and Sónia Paixão for introducing me to the spinal cord and adult brain world, and for the effort to teach me in Italian the related techniques. My special thanks go to Ilona Kadow, Archana Mishra, Dani del Toro, Irina Dudanova, Sónia Paixão, Falko Hampel, Laura Loschek, Alessandro Filosa, Thomas Gaitanos, Pontus Klein, Daniel Nagel and Jingyi Gong for their support, their input, for teaching me several techniques and for all the critical discussions. I would like to thank Louise Gaitanos, Pilar Alcalá and Jana Lindner for their help and for keeping the lab an organized and functional place; Daria Marinescu, Stephanie Krinner and Diana Haba- Schneider for helping me handle the mouse colony. I wish to thank Irina Dudanova and Louise Gaitanos for critically reading my thesis. I would also like to acknowledge the IMPRS coordination office, for their continuous support. Finally, I would like to thank all the present and past members of the Klein lab for the nice and stimulating working atmosphere. My sincerest thanks go to Børk, for being the patient target of my quotidian sarcasm; to Dr. Leidenschaft for being my personal driver and my PR; to God and his wife, for introducing me to the joy of Pizza Hawaii; to Fragile Flower for constantly improving his six pack; to Don Falcone, for showing me that is possible to be happy every single day of your PhD; to the Hulk, for competing with me in G&T consumption; to BamBamDonia, for being such beautiful people; to the Belgian for rocking the dance floor; and to Pizzeria Europa, for feeding me and Dr. Leidenschaft every Thursday night. Finally I would like to thank all the friends with whom I shared happy, sad, complaining, funny or silly moments over the last few years.

Table of Contents

Table of Contents

Abbreviations……...... VI List of Figures……...... X Abstract……………...... XII 1. Introduction……...... 1 1.1. Receptor Tyrosine Kinases: an overview ...... 2 1.1.1. /Trk receptor signaling ...... 3 1.1.1.1. Neurotrophin/Trk signaling for neuron survival ...... 6 1.1.1.2. Other neurotrophin/Trk functions in the central and peripheral ...... 8 1.1.2. Ret/GDNF signaling ...... 10 1.1.2.1. GDNF, Ret and GFRα1: mouse models ...... 12 1.1.2.2. GDNF, Ret and GFRα1 functions in the central and peripheral nervous systems ...... 13 1.1.2.3. GDNF and Ret can signal independently of each other ...... 15 1.1.3. Eph/ signaling ...... 16 1.1.3.1. Distinctive features of Eph signaling ...... 19 1.1.3.2. Eph receptor functions during embryonic development and in adulthood22 1.2. Neuron development: Axon growth and guidance ...... 27 1.2.1. Trigeminal neurons ...... 28 1.2.2. Motor neurons of the lateral motor columns ...... 31 1.3. Intrinsic mechanisms to regulate RTKs signaling ...... 34 1.3.1. Keeping the phosphotyrosine balance: RPTPs versus RTKs ...... 36 1.3.1.1. PTPRO regulation of Trk and Eph receptors ...... 38 1.3.2. Shedding regulates receptor expression and signaling ...... 39 1.3.3. Cooperation of guidance cues and receptor cross-talk ...... 42 1.4. Purpose of thesis project ...... 44 2. Results………………… ...... 45 2.1. PTPRO’s role during development ...... 45 2.1.1. PTPRO’s developmental expression pattern ...... 45 2.1.2. E11.5 and E12.5 PTPRO-/- embryos have bigger and more complex TG arbors...... 51

I

Table of Contents

2.1.3. Cultured E12.5 PTPRO-/- TG neurons display increased sensitivity to BDNF and GDNF, but not NGF ...... 54 2.1.4. Cultured P1 PTPRO-/- TG neurons do not display increased sensitivity to BDNF and GDNF… ...... 58 2.1.5. The exuberant growth and branching observed in PTPRO-/- embryos and neurons are not due to alterations in cell fate or survival...... 60 2.1.6. PTPRO regulates TrkB and Ret signaling ...... 63 2.1.7. PTPRO does not regulate Eph receptors in the developing TG ganglion ...... 68 2.1.8. PTPRO is dispensable as Eph-regulator in LMC axon guidance ...... 69 2.1.9. PTPRO is not required for retinotectal mapping in mouse ...... 72 2.1.10. The chick but not the mouse isoform of PTPRO can dephosphorylate EphA4...... 73 2.2. Role of EphA4 cleavage during development ...... 74 2.2.1. EphA4 is cleaved in Hela and HEK293 cells, independently of ligand stimulation...... 74 2.2.2. EphA4 shedding during embryonic development is temporally and spatially regulated ...... 77 2.2.3. Identification of the EphA4 cleavage site ...... 80 2.2.4. In vitro characterization of the EphA4CR mutant ...... 82 2.2.5. Generation of the EphA4CR knock-in mouse ...... 84 2.2.6. EphA4 expression in EphA4CR/CR embryos ...... 86

2.2.7. EphA4 shedding is required for LMCL neuron axon guidance ...... 89 2.2.8. EphA4 shedding is dispensable for dorsal funiculus and anterior commissure formation ...... 91 2.3. Receptor cross-talk during development ...... 93 2.3.1. EphA4 and Ret do not interact in LMC neurons ...... 93 2.3.2. EphA4 signaling is not impaired in Ret-/- mice ...... 96 2.3.3. GDNF and ephrinAs cooperate in Motor Axon Turning ...... 98 3. Discussion………...... 99 3.1. Roles of RPTPs during development ...... 100 3.1.1. Regulation and specificity of the phosphatase activity ...... 101 3.1.2. Non cell-autonomous role of PTPRO ...... 103 3.1.3. PTPRO as a potential therapeutic target ...... 105 3.2. How does receptor cleavage regulate axon guidance decisions? ...... 106 CR/CR 3.2.1. Potential molecular mechanisms leading to LMCL misguidance in EphA4 embryos ...... 109

II

Table of Contents

3.2.2. What triggers EphA4 cleavage? ...... 112 3.2.3. EphA4 cleavage in neurodegenerative diseases ...... 113 3.3. Guidance cue integration ...... 115 3.3.1. Additive and non-additive effects of guidance cues ...... 115 4. Materials and Methods ...... 117 4.1. Chemicals and drugs ...... 117 4.2. Reagents……...... 117 4.2.1. Plasmids…… ...... 117 4.2.2. Oligonucleotides ...... 118 4.2.3. Cloning primers ...... 118 4.2.4. Genotyping primers ...... 119 4.2.5. Primary antibodies ...... 120 4.2.6. Secondary antibodies ...... 121 4.2.7. Cell lines…… ...... 121 4.2.8. Media………...... 122 4.2.8.1. Luria-Bertani (LB) medium ...... 122 4.2.8.2. LB plates…...... 122 4.2.8.3. Cell culture media ...... 122 4.2.9. Primary culture reagents ...... 122 4.2.10. Primary culture media ...... 123 4.2.11. Buffers and Solutions ...... 124 4.2.12. Mouse lines ...... 128 4.3. Methods………...... 129 4.3.1. Molecular Biology ...... 129 4.3.1.1. Preparation of plasmid DNA ...... 129 4.3.1.2. Transformation of competent E. coli by electroporation ...... 129 4.3.1.3. Site-direct mutagenesis ...... 129 4.3.1.4. TOPO cloning ...... 130 4.3.1.5. Tail DNA preparation and genotyping using PCR ...... 130 4.3.1.6. Agarose gel electrophoresis ...... 131 4.3.2. Cell culture… ...... 132 4.3.2.1. Propagation, thawing and freezing of mammalian cells ...... 132 4.3.2.2. Transfection of cell lines using Lipofectamine ...... 132 4.3.2.3. Primary culture of dissociated mouse trigeminal neurons ...... 133

III

Table of Contents

4.3.2.4. Explant of trigeminal neurons ...... 134 4.3.2.5. Primary culture of dissociated mouse motor neurons ...... 134 4.3.2.6. Explant culture of mouse motor neurons ...... 135 4.3.2.7. Primary culture of dissociated cortical neurons ...... 136 4.3.3. Biochemistry ...... 136 4.3.3.1. Cell lysis and immunoprecipitation of proteins ...... 136 4.3.3.2. Immunoblotting ...... 137 4.3.4. Immunofluorescence ...... 137 4.3.5. Mouse work ...... 138 4.3.6. Histology……...... 138 4.3.6.1. Cryostat sections ...... 138 4.3.6.2. Whole mount Neurofilament staining ...... 139 4.3.6.3. Staining of tissue sections ...... 139 4.3.6.4. Labeling of explant cultures and dissociated motor neurons ...... 140 4.3.6.5. retrograde tracings ...... 140 4.3.7. Generating EphA4CR/CR knock-in mouse ...... 141 4.3.7.1. Cloning…...... 141 4.3.7.2. ES cells culture and DNA electroporation ...... 141 4.3.7.3. DNA extraction ...... 142 4.3.7.4. Southern Blot ...... 142 5. Bibliography…… ...... 144 6.Curriculum vitae ...... 173

IV

V

Abbreviations

Abbreviations

°C degree Celsius aAC anterior AC AC anterior commissure ADAM A disintegrin and metalloprotease APP amyloid precursor protein APS ammonium persulfate ARTN Bax Bcl-2 associated X protein BDNF brain derived neurotrophic factor BSA bovine serum albumin C cysteine-rich domain Cad cadherin-like domain cAMP cyclic monophosphate cGMP cyclic guanosine monophosphate CGRP calcitonin gene related peptide CIN commissural CM cutaneous maximus CMB cell mask blue CNS CNTF ciliary neurotrophic factor CPG central pattern generator CPSG chondroitin sulfate proteoglycan CR cysteine-rich cluster CST corticospinal tract DCC deleted in colorectal carcinoma DF dorsal funiculus DiI 1,1'-dioctadecyl-3,3,3'3'-tetramethylindocarbocyanine perchlorate DIV days in vitro DMEM Dulbecco’s modified Eagle's medium DMSO dimethyl sulfoxide DNA deoxyribonucleic acid dNTP deoxyribonucleotide Dok docking protein DREZ dorsal root entry zone DRG dorsal root ganglia DTT dithiothreitol E embryonic day ECD extracellular domain

VI

Abbreviations

ECL enhanced chemiluminescence EDTA ethylenediaminetetraacetic acid ENC enteric Eph -producing hepatoma EphA4CR EphA4 cleavage resistant EphA4-GFP EphA4 with the intracellular domain replaced by GFP EphA4KD EphA4 kinase-dead EphA4Δ15 EphA4 lacking the external juxtamembrane region EphA4ΔFN3 EphA4 lacking the fibronectin domains EphA4ΔLBD EphA4 lacking the ligand binding domain EphA4ΔN EphA4 lacking the extracellular domain Ephexin Eph interacting exchange protein ERK extracellular signal-regulated kinase ES embryonic stem FBS fetal bovine serum FLRT fibronectin-and-leucine-rich-transmembrane protein FMTC familial medullary thyroid carcinoma FN fibronectin like domain FRS2 fibroblast receptor substrate 2 Gab Grb-associated-binding protein GAP GTPase-activating protein GDNF glial cell line-derived neurotrophic factor GEF guanine exchange factor GFL GDNF-family ligand GFP green fluorescent protein GFRα GDNF family receptors alpha protein GPCR G-protein coupled receptor GPI glycosylphosphatidylinositol Grb -bound protein GTP guanosine-5'-triphosphate HBSS Hank's balanced salt solution HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HGF HRP horse radish peroxidase ICD intracellular domain Ig immunoglobulin-like Ig-CAMs immunoglobulin family cell adhesion molecules INL inner nuclear layer IP3 phosphatidylinositol-(1,4,5) triphosphate IRS receptor substrate

VII

Abbreviations

JNK c-Jun N-terminal kinase kDa kilo Dalton LAR leukocyte common antigen-related LB Luria-Bertani LD latissimus dorsi LMC lateral motor column

LMCL lateral LMC

LMCM medial LMC LRR leucine-rich repeat LTD long term depression LTP long term potentiation MAPK mitogen activated MEN2 multiple endocrine neoplasia type 2 MMP matrix metalloprotease mRNA messenger RNA NCS newborn calf serum NF-κB nuclear factor kappa-light-chain-enhancer of activated B cells NGF NRTN NT-3 neurotrophin-3 NT-4/5 neurotrophin-4/5 ONL outer nuclear layer P post-natal p140NCAM p140 neural cell adhesion molecule p75NTR neurotrophin receptor p75 pAC posterior AC PAGE polyacrylamide –gel –electrophoresis PBM PDZ binding motif PBS phosphate buffered saline PCR polymerase chain raction PI3K phosphoinositide-3-kinase PLCγ gamma PS presenilin PSPN PTB phospho-tyrosine binding PTP protein tyrosine phosphatase PTPRO receptor protein tyrosine phosphatases type O pTyr phosphotyrosine pY1062 phosphotyrosine1062 RBD receptor binding domain

VIII

Abbreviations

RD rhodamine dextran Ret rearranged during transfection RGC retinal ganglion cell RNA ribonucleic acid Robo roundabout RPTP receptor protein tyrosine phosphatases RTK receptor tyrosine kinase RTK room temperature SAM sterile-α-motif SC superior colliculus SCF SCG superior cervical ganglia SDS sodium dodecyl sulfate SEM standard error of the mean Sema semaphorin SH2 src homology 2 Shc Src homologous and collagen-like SHH sonic hedgehog SP substance P STAT signal transducer and activator of transcription TCL total cell lysates TEMED tetramethylethylenediamine TG trigeminal ganglion TGF-β transforming growth factor-β Tris tris(hydroxymethyl)aminomethane Trk tropomyosin-related kinase TZ termination zone VEGF vascular endothelial growth factor Wnt wingless integration Wt wild-type µ Micro

IX

List of Figures

List of Figures

Figure 1-1 Structure of Trk receptors and p75 Figure 1-2 BDNF/TrkB signaling cascade Figure 1-3 Structure of Ret, GFRαs and GFLs Figure 1-4 GDNF/GFRα1/Ret signaling complex Figure 1-5 Structure of Eph receptor and Figure 1-6 Eph/ephrin forward signaling Figure 1-7 Eph/ephrin reverse signaling Figure 1-8 EphA/ephrinA signaling in the retinotopic mapping Figure 1-9 EphB/ephrinB signaling in retinotopic mapping Figure 1-10 Sensory neuron guidance in the trigeminal ganglion Figure 1-11 LMC guidance in the limb Figure 1-12 Axon guidance at the spinal cord midline Figure 2-1 Specificity of the anti-PTPRO antibody Figure 2-2 PTPRO expression pattern in LMC neurons Figure 2-3 PTPRO expression pattern during TG development Figure 2-4 PTPRO expression in a subset of TG neurons Figure 2-5 PTPRO expression in a subset of DRG neurons Figure 2-6 PTPRO expression at the spinal cord midline Figure 2-7 PTPRO expression in the retina Figure 2-8 E11.5 PTPRO-/- embryos have a more complex ophthalmic arbor Figure 2-9 E12.5 PTPRO-/- embryos show exuberant arborization of the ophthalmic branch of the TG nerve Figure 2-10 E12.5 PTPRO-/- embryos show defasciculation of the maxillary branch Figure 2-11 E12.5 PTPRO-/- TG neurons are more sensitive to BDNF and GDNF Figure 2-12 E12.5 PTPRO-/- TG neurons are more sensitive to BDNF and GDNF, but not NGF stimulation Figure 2-13 P1 PTPRO-/- TG neurons do not show increased sensitivity to and GDNF Figure 2-14 PTPRO-/- embryos do not have defects in TG neuron differentiation Figure 2-15 PTPRO-/- embryos do not have defects in DRG neuron differentiation Figure 2-16 Loss of TrkA+ and TrkC+ neurons in newborn PTPRO-/- mice Figure 2-17 Regulation of TrkB signaling by PTPRO in transfected cells Figure 2-18 Regulation of Ret51 signaling by PTPRO in transfected cells Figure 2-19 PTPRO does not regulate TrkB and Ret 51 surface expression Figure 2-20 PTPRO-/- TG explants do not show increased sensitivity to ephrinAs Figure 2-21 Neurofilament staining on whole-mount PTPRO-/- embryos does not show any guidance defects Figure 2-22 PTPRO-/- motor neurons are not more sensitive toward ephrin stimulation

X

List of Figures

Figure 2-23 PTPRO-/- mice do not show misguidance or aberrant branching in the retinocollicular map Figure 2-24 The chick but not the mouse isoform of PTPRO can dephosphorylate EphA4 Figure 2-25 EphA4 is cleaved in Hela and HEK293 cells Figure 2-26 EphA4 cleavage is proportional to EphA4 expression levels Figure 2-27 EphA4 cleavage is independent of ligand stimulation Figure 2-28 EphA4 is cleaved in E16.5 cortical neurons and cleavage regulates receptor levels in culture Figure 2-29 EphA4 is cleaved in vivo Figure 2-30 EphA4 cleavage is spatially and temporally regulated Figure 2-31 EphA4 cleavage has a peak between E12.5 and E15.5 Figure 2-32 Eph receptor cleavage Figure 2-33 EphA4Δ15 is still cleaved with low efficiency Figure 2-34 EphA4CR is cleavage resistant Figure 2-35 EphA4CR is expressed on the cell surface and it is phosphorylated upon ephrinA5 stimulation Figure 2-36 EphA4CR shows increased trans-endocytosis into ephrin expressing cells Figure 2-37 Generation of the EphA4CR knock-in mouse Figure 2-38 EphA4CR is sufficient to abolish receptor cleavage in vivo Figure 2-39 EphA4CR/CR has increased levels of EphA4 full-length protein Figure 2-40 In EphA4CR/CR embryos full-length EphA4 is up-regulated in the hindlimb mesenchyme but not on motor axons Figure 2-41 In EphA4CR/CR embryos full-length EphA4 is up-regulated in the dorsal spinal cord but not on motor neurons Figure 2-42 Hindlimb retrograde tracings show misguidance of LMCL neurons in EphA4CR/CR embryos Figure 2-43 Dorsal funiculus morphology and anterior commissure formation are not affected in EphA4CR/CR mice Figure 2-44 Characterization of dissociated LMC cultures Figure 2-45 Specificity of Ret and EphA4 antibodies Figure 2-46 Ret and EphA4 do not directly interact in motor axons Figure 2-47 EphA4 phosphorylation and shedding are not altered in E12.5 Ret-/- embryos Figure 2-48 EphA4-induced collapse is not affected in Ret-/- embryos Figure 2-49 Cooperation between GDNF and ephrinA5 in motor axon turning Figure 3-1 Models for the regulation of RPTP phosphatase activity

Figure 3-2 Hypothetical molecular mechanisms leading to LMCL misguidance in EphA4CR/CR mice

XI

Abstract

Abstract

Axons are equipped with an exploratory tip, the growth cone, to navigate and sense

the cues presented by the surrounding environment. Several families of ligands are

present along the axonal pathways, while their receptors are expressed on the growth

cone and allow different axons to follow a great variety of trajectories. However, the

number of molecules involved could be considered relatively small if compared to the

diversity of trajectories, speed of growth and arborization patterns present in developed

organisms. The fine tuning and the integration of different guidance cues represent good

mechanisms to amplify and diversify the outputs of a relatively small number of

ligand/receptor systems. The molecular players taking part in the modulation and

integration of different signaling are not yet fully elucidated. In this study I focused on

three intrinsic mechanisms to modulate receptor tyrosine kinase signaling:

dephosphorylation by receptor protein tyrosine phosphatases (RPTPs), receptor cleavage

and receptor cross-talk.

First, I analyzed TrkB, Ret and Eph receptor interaction with RPTP type O (PTPRO)

in trigeminal and motor neurons. PTPRO is expressed mainly in TrkB+ and Ret+ mechanoreceptors within the TG during embryogenesis. In PTPRO mutant mice, the maxillary and ophthalmic branches of the trigeminal ganglion grow more complex arbors than in littermate controls. Cultured PTPRO-/- TG neurons display enhanced axonal

outgrowth and branching in response to BDNF and GDNF compared to control neurons,

indicating that PTPRO negatively controls the activity of BDNF/TrkB and GDNF/Ret

signaling. Mouse PTPRO fails to regulate Eph signaling in retinocollicular development,

XII

Abstract

in hindlimb motor axon guidance, and in transfected heterologous cells, suggesting that

chick and mouse PTPRO have different substrate specificities.

On a second approach to identify intrinsic mechanisms to regulate receptor signaling,

I analyzed how receptor cleavage regulates EphA4 signaling during development. Upon

characterizing EphA4 cleavage in vitro, I generated a knock-in mouse carrying a

mutation that made the EphA4 receptor cleavage resistant (EphA4CR). Abolishing EphA4

cleavage led to an increased expression of the full-length protein in hindlimb mesenchyme and in dorsal spinal cord, but not on motor neuron soma or axons.

CR Moreover, in EphA4 embryos, LMCL neurons were aberrantly rerouted to the ventral

mesenchyme, similarly to the guidance defects observed in EphA4-/- embryos.

Interestingly, two other phenotypes present in EphA4-/- mice, the shallowing of the dorsal

funiculus and the loss of the anterior commissure, were not present in EphA4CR mice,

suggesting that cleavage is only required for certain EphA4 functions.

Finally, I studied, in collaboration with Dr. Irina Dudanova, the molecular

mechanisms underlying EphA4 and Ret cooperation in motor axon guidance at the sciatic

plexus. We demonstrated that the two signaling systems act in parallel and independently

to guide LMCL axons in the dorsal mesenchyme of the hindlimb. When presented as

opposing gradients, GDNF and ephrinAs cooperated and triggered a stronger turning

response, suggesting that Ret and EphA4 exert different effects on the same growth cone.

The in vitro results were consistent with the in vivo expression of the two proteins, where

GDNF expressed dorsally to the choice point attracts LMCL axons, and ephrinAs

expressed ventrally repel them. This represents the first example of two opposing cues

XIII

Abstract acting in an additive manner to promote the same guidance choice at an intermediate target.

Taken together these data provide new insights in understanding the regulation of receptor signaling by modulatory proteins or by other receptors.

XIV

Introduction

1. Introduction

During embryonic development neurons need to find their appropriate synaptic targets among many possible. Each axon terminates with an exploratory tip, the growth cone, which is equipped with several receptors to sense different cues in the surrounding environment. These cues can be either membrane-bound or soluble, and can provide trophic or tropic support. Several families of receptors/ligands are expressed on growth cones and in their target tissues and allow different axons to follow a great variety of trajectories. Neurons receive support for their outgrowth, branching and survival from and guidance direction from several families of axon guidance molecules [1].

In the last decades, four conserved families of axon guidance molecules have been identified: ephrins, netrins, semaphorins and Slits. In addition to these well characterized families, further guidance factors have more recently been described, e.g. morphogens such as Wnts and sonic hedgehog (SHH), growth factors such as hepatocyte growth factor (HGF), glial cell line-derived neurotrophic factor (GDNF), immunoglobulin family cell adhesion molecules (Ig-CAMs), and protocadherin family (reviewed in [2]).

However, the numbers of ligand/receptor systems can be considered relatively small if compared to the complexity of the nervous system. Nonetheless, regulation and integration of guidance cues may represent mechanisms by which only a few molecules are sufficient to ensure the correct formation of a great variety of structures in the nervous system (as well as in other tissues). Work over the past years has identified several means of yielding diverse outcomes from the same ligand/receptor system: firstly, the controlled regulation of the molecule’s expression (by alternative splicing,

1

Introduction microRNAs, etc.); secondly, the intrinsic (neuron type-specific) or extrinsic regulation of the pathways; lastly, the interaction with other receptors [3].

1.1. Receptor Tyrosine Kinases: an overview

Cells express on their surface a plethora of receptors to transduce a great variety of extracellular stimuli. There are three main classes of receptors: G-protein coupled receptors (GPCRs), ion channels and -coupled receptors. The latter can either act as an enzyme upon ligand binding or be associated with an enzyme. Among the enzyme- coupled receptors having their own catalytic activity, the most prominent family is the receptor tyrosine kinase family (RTKs) [4].

In humans there are 20 subfamilies of RTKs, which share similar structures. The mechanism of activation and the downstream pathways are conserved from nematode to humans. that affect RTK activity, abundance, cellular localization or tissue expression are associated with numerous diseases, including inflammation, cancer, diabetes, and arteriosclerosis [5].

Generally, RTKs are activated by dimerization and act on common downstream pathways: mitogen activated protein kinase (MAPK), Akt, phospholipase C gamma

(PLCγ) and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB). The first substrates of the kinase activity are the tyrosine residues present on the receptor, which then act as docking sites for adaptor proteins containing Src Homology 2 (SH2) and Phospho Tyrosine Binding (PTB) domains. In the absence of a ligand, kinase activity is often blocked by an auto-inhibitory mechanism, which can vary among different

RTKs. For example, in the the auto-inhibitory tyrosine is in the kinase loop, whereas in the MuSK receptor the auto-inhibitory tyrosines are located in the

2

Introduction juxtamembrane region. Moreover, when not bound to their ligands, receptors can be kept in a dephosphorylated state by interaction with protein tyrosine phosphatases (PTPs).

Pharmacological blockade of PTPs results in a general increase of RTK activation [5].

Once the receptor has been activated, it can receive positive and negative feedbacks, which can modulate the strength and the duration of the signaling output [5]. Amongst others, two families of transmembrane proteins have been characterized as RTK regulators: the receptor protein tyrosine phosphatase (RPTP) family and the LIG family of leucine-rich repeat (LRR) and immunoglobulin proteins.

In the following paragraphs I will focus on three RTK subfamilies that play a well- established role in neuron growth and guidance: Trk, Ret and Eph receptors.

1.1.1. Neurotrophin/Trk receptor signaling

The neurotrophin family, in mammals, has four members: nerve growth factor (NGF), brain derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3) and neurotrophin-4/5

(NT-4/5). Neurotrophins act as dimers and are secreted as precursors (pro-neurotrophins).

Pro-neurotrophins can be cleaved intracellularly (in the trans-Golgi network) by furin and other pro-hormone convertases, or extracellularly by plasmin. Neurotrophins bind to two classes of receptors: the tropomyosin-related kinase (Trk) receptor family and neurotrophin receptor p75 (p75NTR), a member of the tumor necrosis receptor superfamily. p75NTR is a common receptor for all neurotrophins, and although it lacks a catalytic domain, it regulates neuronal survival and differentiation through interaction with other receptors. p75NTR can also act as a co-receptor for the Trk receptors, increasing their affinity for the ligand [6, 7].

3

Introduction

In mammals, the Trk receptor family has only three members: TrkA, TrkB and TrkC.

Each receptor is characterized by the presence of two cysteine-rich clusters, three

leucine-rich repeats and two immunoglobulin-like domains in the extracellular region, a

transmembrane domain and an intracellular kinase domain (Figure 1-1). The membrane-

proximal immunoglobulin domain has been described as important for the binding of

neurotrophins [8]. Trk receptors undergo alternative splicing generating several isoforms,

which can either differ by a few amino acids within or around the immunoglobulin

domain, or be truncated versions of the receptors, lacking portions of the intracellular

domain. Differences in the immunoglobulin domain modify the affinity of Trk receptors

to specific neurotrophins, generally to the non-preferred ligands [9, 10]. The truncated receptors have different functions than their full-length counterparts: they can either initiate their own signaling cascade or act as dominant negative regulators of Trk signaling [11, 12].

NGF BDNF NT-4/5 NT3

CR1 LRR1-3 CR1 CR2 CR2 CR3 Ig1 CR4 Ig2

Tyrosine Kinase

p75 TrkA TrkB TrkC Figure 1-1. Structure of Trk receptors and p75 Schematic drawing of Trk receptors. The Trk extracellular region contains two cysteine-rich clusters (C1- 2), three leucine-rich repeats (LRR1-3) and two immunoglobulin-like domains (Ig1-2). The intracellular region has a tyrosine kinase domain. p75 has four cysteine-rich clusters (CR1-4) and an intracellular domain lacking kinase activity. NGF is TrkA ligand, BDNF and NT-4/5 are TrkB ligands, and NT-3 is the ligand for TrkC (black arrows). p75 binds all neurotrophins with low affinity (grey arrows).

4

Introduction

TrkA, TrkB and TrkC bind with high affinity to NGF, BDNF and NT-3, respectively.

TrkB can also bind NT-4/5 (Figure 1-1). Upon ligand binding these receptors dimerize,

trans-phosphorylate the tyrosine residues in their intracellular domains, and activate

several signaling pathways. In vertebrates, Trk receptors have 10 conserved tyrosine

residues that can be phosphorylated upon ligand binding. Three of these tyrosines are

present in the autoregulatory loop of the kinase domain, thereby controlling receptor

activation [13].

STAT3 SOS Y484 Shc Ras P Grb2 BDNF Frs2 c-Abl

Crk PI3K Raf Y670 P Y674 P PDK1 MEK C3G Y675 P TrkB Akt Erk

2+ Y785 P IP3 Ca PLCγ DAG PKC

Figure 1-2. BDNF/TrkB signaling cascade Schematic drawing of TrkB signaling. Upon BDNF binding TrkB forms dimers and several of its intracellular tyrosines become autophosphorylated. The phospho-residues act as docking sites for few adaptor proteins, which activate several downstream pathways, like MAPK/ERK, PI3K, PLCγ.

In TrkA the two main phospho-tyrosines are tyrosine 490 (tyrosine 484 in TrkB) and

tyrosine 785 [14]. Tyrosine 490 acts as a docking site for Src homologous and collagen-

like (Shc) and receptor substrate 2 (FRS2), and tyrosine 785 for

PLCγ. Shc triggers the transient activation of Ras, which then starts the phosphoinositide-

3-kinase (PI3K) and MAPK/ERK signaling pathways; FRS2 recruits Crk, which binds

the guanine nucleotide exchange factor, C3G [15-17]. Recruitment and phosphorylation

5

Introduction of PLCγ leads to formation of phosphatidylinositol-(1,4,5) triphosphate (IP3), which stimulates the release of calcium from intracellular storage compartments, and diacyl glycerol (DAG), which activates protein kinase C (PKC) [8]. Although tyrosines 490 and

785 are the main phosphorylation sites, knock-in mice in which these tyrosines have been converted to phenylalanine do not show major abnormalities, suggesting that there is a redundancy of phospho-tyrosines that can start the downstream signaling pathways [18-

20]. Finally, the tyrosines in the autoregulatory loop can recruit growth factor receptor- bound protein 2 (Grb2) [21, 22], and c-Abl can also bind to non-phosphotyrosine residues

[23, 24] (Figure 1-2).

1.1.1.1. Neurotrophin/Trk signaling for neuron survival

According to the neurotrophic factor hypothesis, between embryonic day 13 (E13) and 18 (E18) neurons generated in excess during development undergo programmed cell death, because they compete for limited amount of neurotrophic factors present in the target tissues [25, 26]. Genetic ablation of neurotrophin and Trk genes in most of the cases, with the exception of NT-4/5, affects mouse viability and the survival of several populations of peripheral neurons [13]. Although in vitro neurotrophins promote survival of several populations of neurons, in vivo their role seems to be restricted to specific populations [8].

NGF and TrkA knockout mice display loss of neurons in superior cervical ganglia

(SCG), dorsal root ganglia (DRG) and trigeminal ganglia (TG). In the DRG there is a loss of calcitonin gene related peptide positive (CGRP+), IB4 positive (IB4+) and substance P positive (SP+) neurons, and in the spinal cord, projections to lamina I and II (nociceptive fibers) are lost. As a consequence of this loss of nociceptive neurons, knockout mice are

6

Introduction

less sensitive to pain. Moreover, these mutant mice have a reduced number of low-

threshold mechanoreceptor [27-30]. In the central nervous system (CNS), TrkA and NGF knockouts show loss of cholinergic projections, although the number of neurons is not affected [30].

TrkB and BDNF knockout mice display loss of SCG, TG, vestibular, nodose, trigeminal mesencephalic nucleus and DRG neurons. The DRG neurons lost in these knockouts are a subset of the cutaneous mechanoreceptors. NT-4/5 knockout mice have a reduced number of nodose and geniculate neurons, and this phenotype is enhanced in NT-

4/5-/-;BDNF-/- mice [31-35].

Based on their expression patterns, TrkC and NT-3 have been associated with

neurons responsible for . Consistent with this observation, mutant mice for

either the ligand or the receptor are impaired in movements and have abnormal postures.

NT-3 and TrkC mice display loss of neurons in the SCG, in the TG, in the nodose

ganglion, in the cochlear ganglion and in DRGs. Sensory projections connecting to motor

pools in the spinal cord (Ia projections, proprioceptive axons) are missing. Moreover, in

these mice Golgi tendon organs, muscle spindles and sensory peripheral innervation are

absent. NT-3 mutant mice show a more severe phenotype than the TrkC knockouts,

suggesting that NT-3 may have additional receptors [36-41].

Interestingly, a recent paper showed that TrkA and TrkC, but not TrkB are able to

signal independently of neurotrophin binding. Over-expression of these receptors is

sufficient to trigger cell death in absence of the ligand, and if NGF or NT-3 are added to

the neurons, cell death is rescued [42]. This data further prove the hypothesis that TrkA

and TrkC act as dependence receptors. Dependence receptors are receptors able to initiate

7

Introduction

two signaling cascades: one in the presence of ligand, leading to survival, differentiation

or migration; and another one in the absence of the ligand, which triggers or amplifies

signaling, leading to programmed cell death [43].

1.1.1.2. Other neurotrophin/Trk functions in the central and peripheral nervous

system

In addition to their well-established roles in neuron survival, neurotrophins and Trk receptors have been implicated in differentiation, modulation of axonal and dendrite

outgrowth and guidance, and in the regulation of [13].

In vivo, it has been possible to uncover additional functions of neurotrophin signaling

only upon crossing neurotrophin and neurotrophin receptor mutants with Bcl-2 associated

X protein (Bax) knockouts. Removing Bax prevents , and allows the uncoupling

of neurotrophin effects on survival from those on specification. TrkA/Bax and NGF/Bax

double knockouts show a milder loss of neurons compared to TrkA or NGF single

knockouts. In NGF-/-, TrkA+ neurons are unable to differentiate into CGRP+, Ret+ and

SP+ neurons [44]. This in vivo data are supported by the ability of NGF to induce neuropeptide expression in cultured embryonic DRG neurons [44].

A role for neurotrophins as guidance molecules has been speculated since the discovery of NGF-induced outgrowth in cultures [45]. All neurotrophins trigger neurite outgrowth in embryonic sensory neuron cultures [46, 47]. The in vivo relevance

of TrkA signaling in supporting neurite outgrowth was assessed, as described before for

TrkA role in differentiation, in TrkA/Bax knockouts. In these mice spinal cord innervation

is unaffected, but cutaneous innervation is disrupted, suggesting that NGF/TrkA signaling

is required for peripheral innervation, and the absence of projections in the spinal cord of

8

Introduction

NGF or TrkA knockouts is secondary to cell death [44]. Mice over-expressing NGF or

BDNF in the dermis provide further evidence for a role of neurotrophins in regulating

peripheral innervation. since these transgenic mice display hyper-innervation of the

whisker pad and the dermis [48]. In addition to their trophic functions, neurotrophins can

act as attractive guidance cues for mouse DRG neurons and Xenopus spinal neurons

when presented in a gradient [47, 49]. Surprisingly, neurotrophins can also act as

chemorepellents when cyclic adenosine monophosphate (cAMP) and cyclic guanosine

monophosphate (cGMP) are inhibited [49]. Although in vitro neurotrophins are able to

steer the growth cones of several types of neurons, in vivo data support a role for

neurotrophins in regulating outgrowth, branching and target innervation of several

neuronal populations, but not axon pathfinding [50].

A role for neurotrophins in regulating synaptic plasticity is shown by several lines of

evidence, including the regulation of their secretion by neuronal activity and their ability

to potentiate synaptic transmission [51]. Neurotrophins are also able to induce structural

changes, i.e. regulate the size of dendritic arbors of pyramidal neurons, and to enhance short- and long-term synaptic transmission. BDNF and TrkB mutants show severe impairment of LTP, although basal synaptic transmission is not affected [52, 53].

Consistently with impairment in LTP, neurotrophin mutant mice have several behavioral abnormalities when performing tasks to assess their ability to learn or memorize. The conditional knockout approach allowed the investigation of behavioral defects in mice where TrkB was specifically removed from the : this resulted in abnormal

memory acquisition and consolidation in hippocampus-dependent learning tasks [53].

9

Introduction

Similarly to BDNF-/- mice, reduction of NGF levels (NGF heterozygous mice) caused impairment in the formation and retention of memory [54].

1.1.2. Ret/GDNF signaling

GDNF-family ligands (GFLs) belong to the transforming growth factor-β (TGF-β) super-family. They are characterized by the presence of six cysteine residues regularly spaced to form 3 disulfide bonds (cysteine knot). They are secreted as precursors

(preproGFLs), and after being activated by proteolytic cleavage, function as homodimers

[55]. The four GFLs - GDNF, neurturin (NRTN), artemin (ARTN) and persephin (PSPN)

- signal via Ret, a transmembrane receptor tyrosine kinase, and one of the four GPI-

anchored GDNF family receptors alpha proteins (GFRα1-4) [56] (Figure1-3).

GDNF NRTN ARTN PSPN

Ret Cad1 Cad2 CR1 Cad3 CR2 Cad4 CR3 CR1

GFRα1GFRα2GFRα3 GFRα4 Tyrosine Kinase

Figure 1-3. Structure of Ret, GFRαs and GFLs Schematic drawing of Ret receptors, GFRα co-receptors and GFLs. The extracellular region of Ret contains four cadherin-like domains (Cad1-4) and one cysteine-rich domain (C1). The intracellular region has a large intercalated tyrosine kinase domain. GFRα1, GFRα2 and GFRα3 have three cysteine-rich clusters (CR1-3), whereas GFRα4 has only two. GFLs act as dimers. All GFLs bind to Ret, but using different co- recept